54. On all grounds, therefore, we must say that between conscious, sub-conscious, and unconscious states the difference is only of degree of complication in the neural processes, which by relative preponderance in the consensus determine a relative discrimination. We can only discriminate one thrill at a time; but the neural excitations simultaneously pressing towards a discharge are many; and the conditions which determine now this, and now the other excitation to predominate by its differential pressure, are far beyond any mechanical estimate. I mention this because the advocates of the Reflex Theory maintain that the neural processes are the same whether a sensation be produced or not; and that since the same actions follow the external stimulation whether sensation be produced or not, this proves the actions to be purely mechanical. I reply, the neural processes are _not_ the same throughout in the two cases--otherwise the effects would be the same. You might as well say, “Since the explosion of the gun is the same, whether shotted or not, a blank cartridge will kill”; but if you tell me that your gun killed the bird, I declare that the cartridge was _not_ a blank one. Whether the explosion of the gun _also_ produced terror in one bystander, curiosity in a second, and attracted no notice from a third, will be altogether another matter. In like manner the sensory impression which determines a movement may or may not be accompanied or followed by _other_ sentient states; the fact of such movement is evidence of its sentient antecedent; and an external stimulus that will produce _this_ neural process, and this consequent movement, must produce a feeling, although not necessarily a discriminated sensation. Now since, for discrimination, _other_ neural processes must co-operate, we cannot say that in the two cases _all_ the neural processes have been the same throughout; nor because of this difference can we say that the process of the undiscriminated sensation is a mechanical, not a sentient process. In the next problem this point will be argued more fully.